From: Kevin Carhart <kevin@carhart.net>
To: Karl Dahlke <eklhad@comcast.net>
Cc: Edbrowse-dev@lists.the-brannons.com
Subject: Re: [Edbrowse-dev] eb$master
Date: Tue, 19 Sep 2017 01:16:11 -0700 (PDT) [thread overview]
Message-ID: <alpine.LRH.2.03.1709190048430.12633@carhart.net> (raw)
In-Reply-To: <20170819015218.eklhad@comcast.net>
Thank you for doing this, I'm glad you have an idea of what to do. I'm
also worried about the underlying speed issue:
> nasa.gove takes 1 minute 30 to browse, which is already unacceptably
> slow.
Do you know where the bulk of the delay lies? I thought maybe it was
related to the CSS. Doling out styles to elements, maybe inefficiently,
like a cartesian product. But I am actually not sure.
Is it fair to say that the slow pages like nasa.gov are not a regression,
but we now have slow things that weren't supported at all, previously? Is
there a danger of leaving anyone worse off?
> showscripts
These new utilities are very exciting. Also, til today I had never run
the standalone duk even though you had mentioned it. I had only compiled
the library thus far. The standalone duk is very useful in conjunction
with demin, because standalone duk gives you a stack trace for errors,
with a line number for every call! If good line numbers are available you
get very rich information. I had no idea we could find this out.. it's a
new world, at least for find & fix!
On Tue, 19 Sep 2017, Karl Dahlke wrote:
> As per my last post, I now give each window access to the master window through eb$master.
> This window is always there, behind the scenes.
> Even if you unbrowse every file, so it looks like there is no js anywhere in edbrowse, the master window persists.
> If functions have been compiled in the master window then eb$master.compiled will be true.
> We can, in *.js, test for this, and if it is false we can compile a large static function in the master window,
> then link to it from our window.
> I have done this for dumptree(), showscripts(), and the console object.
> These functions are compiled once but available from every window.
> If things don't blow up I'll do the same for most of the other functions,
> especially the monster functions in third.js.
> This makes it a little less cut&paste to put in an updated snapshot of one of these functions however.
> If it says function foo() we have to change that to eb$master.foo = function()
> but other than that it isn't hard, just something we maintainers have to remember to do.
> Check it out, and if you don't object I will gradually transform the other static functions into a compile-once mode.
> I don't bother doing this for the little one or two line functions,
> and certainly no point for the native methods, since a pointer to C takes up less space than a pointer to eb$master.whatever.
>
> Karl Dahlke
>
--------
Kevin Carhart * 415 225 5306 * The Ten Ninety Nihilists
prev parent reply other threads:[~2017-09-19 8:14 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-09-19 5:52 Karl Dahlke
2017-09-19 8:16 ` Kevin Carhart [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=alpine.LRH.2.03.1709190048430.12633@carhart.net \
--to=kevin@carhart.net \
--cc=Edbrowse-dev@lists.the-brannons.com \
--cc=eklhad@comcast.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).